/deep-neural-networks

These are are the different neural networks coded in vanilla Java without using any deep learning framework

Primary LanguageJava

Deep Neural Networks

These are are the different neural networks coded in vanilla Java without using any deep learning framework

This code was written jointly by Felipe Gutierrez (git_id: felipegb94) and Nikhil S. Nakhate (git_id: nikhil-nakhate) as a part of the coursework for CS838.

All the neural networks have early stopping implemented in them.

ANN: This folder contains a dataset along with the source files which is the only argument that is required to run the code. The code would handles the sampling into a train_set, tune_set and test_set

CNN: In addition to the source files, this folder also contains a MakeFile for compilation and a run file that includes the parameters to be passed.

The first three params would be the train_set, tune_set and test_set respectively and the fourth param would be size of the images used.